The medical, social, and economic burdens that Alzheimer's disease (AD) and the prodromal stage of this pathology, mild cognitive impairment (MCI), place on Veterans has stimulated efforts to identify therapeutic targets to modify their progression. Recent evidence suggests that, in addition to vascular abnormalities in AD, a deficiency in endothelium-derived nitric oxide (NO) may contribute to the production and accumulation of AD- related neurotoxins such as amyloid beta (A?). As challenges, which we are uniquely poised to overcome, have deterred the in vivo assessment of cerebral endothelial function and NO bioavailability, it is unclear if these factors are, indeed, attenuated in AD. Therefore, the first aim of this study is to determine if endothelial function is related to AD severity by measuring the change in cerebral blood flow elicited by the intravenous infusion of the NO synthase (NOS) inhibitor NG-monomethyl-L-arginine (L-NMMA) to directly assess cerebral NO bioavailability in patients with AD, MCI, and cognitively-normal age and sex matched controls. We also propose to determine if simpler, non-invasive assessments of peripheral endothelial function are good surrogates for cerebral endothelial function. The second aim of this study will determine the role of endothelial function and NO bioavailability in AD progression by longitudinally assessing and comparing the change in endothelial function (cerebral and peripheral) to the change in cognitive function over the course of 18 months in patients with MCI and age and sex matched controls. Thus, the overall goal of the proposed studies is to better understand the role of the vascular endothelium in AD severity and progression to determine if cerebral endothelial function and NO bioavailability are viable therapeutic targets to limit this debilitating disease.